Proceedings o f The South Africun Sugar Technologists' Associutiorz - April 1968

'THE DESIGN OF BELT CONVEYORS FOR

BULK SUGAR HANDLING
By A. M. GUTHRIE and J. R. PILCHER

Moreland Technical and Engineering

Consultants Ltd.

1. Introduction
Depending on the condition of the sugar and
the quantities to be handled, many different methods
are employed for transporting sugar in bulk, including, inter a1ia:-Screw conveyors;
Grasshoppers;
Belt conveyors;
Bucket: elevators.
Most of the above-mentioned types of conveyors
may be seen in almost any sugar factory, and, as
any sugar engineer will be only too quick to explain,
each method has its own peculiar problems which
suit it to certain applications.
Belt conveyors have their attraction for most
applications because of the small number of wearing parts, which results in low maintenance costs;
and the absence of relative movement between the
sugar and conveyor components with which it is in
contact, which results in minimal product degradation. However, they can be extremely difficult to
install succesqfully. In fact, one well-known sugar
engineer who has a number of belt conveyors
handling his final sugar was heard to comment that
"his belt conveyors were fine, they would take away
as much sugar as you can put on them; the snag
is, they bring most of it back again."
It may be stated that almost anyone could successfully design a belt conveyor to carry a given quantity of sugar, or any other commodity. The secret of
the successful design of a conveying system lies
in the design of the transfers to and from the conveyor, to ensure maximum loading of the belt with
minimum spillage.
Where the quantities of sugar to be handled are
low (less than 50 tons per hour) and conveying
distances are comparatively short, other methods of
conveying sugar are frequently found to be more
suitable despite higher maintenance and operating
costs; but once high conveying rates and long distances become necessary, as at the Sugar Terminal,
there would appear to be no alternative to the belt
conveyor, and under these circumstances the control
of spillage becomes of paramount importance. At
the Sugar Terrninal certain of the conveyors are
designed for a duty in excess of 1,000 tons per hour,
at which duty a spillage of only 0.1% will leave
24 tons of sugar per day to be shovelled off the
floor. This is obviously totally unacceptable, and a
thorough knowledge of the behaviour of sugar at a
belt transfer becomes a prerequisite to designing an
acceptable system.
The question may be asked, "What about the
possibility of pneumatic conveying?" The authors

are not aware of a successful pneumatic conveying

system for raw sugars, particularly where low pol
varieties have to be handled, due to the sticky nature
of the material to be carried, and a pneumatic conveying system for refined sugars would be hazardous
in the extreme due to the possibility of the sugar
dust forming an explosive mixture with the air.
,
2. Selection of Conveyor Components
This subject is dealt with exhaustively in trade
literature, and it is only proposed to mention the
main points to be considered briefly.
Assuming that the rate at which it is desired to
transport the product has been established, it is
then necessary to decide on the belt speed, belt
width, idler type and type of belt in order to
establish the basic parameters of the conveyor
design. All the above variables are to a greater or
lesser degree inter-dependent and the success and
economy of the final design depends on achieving
the optimum balance between them.

2.1 Belt Speed

Let us in the first instance examine the effect of
varying belt speeds. The range of speeds which has
been used in belt conveyor design varies from
approximately 150 feet per minute to as high as
1,500 feet per minute. In general, speeds above
300 to 350 feet should only be contemplated for
high conveying rates where it is necessary to use
belts wider than 30" in order to achieve the
desired material transfer rate. The higher the belt
speed the more difficult it is to load the belt without
spillage. Wider belts alleviate the loading problem
and the highest belt speeds have been used on belts
of 48" and 54" width. This choice of parameters
(i.e. 48" width and 1;500 ft. per minute belt speed)
gives material transfer rates, depending on the density of the material being handled, of 2,000 to 3,000
tons per hour in a deep-troughed belt. Thus, in
general terms, it may be stated that for belts less
than 30" wide belt speeds in excess of 400 ft. per
minute should be avoided, while belt speeds above
1,000 ft. per minute should be avoided on belts less
than 48" wide. I t is axiomatic that provided the
belt can be adequately loaded the carrying capacity
is proportional to the belt speed. Unfortunately, with
narrow belts at high speeds it becomes increasingly
difficult to achieve an adequate degree of loading
without excessive spillage of product at the transfer
points.
2.2 Idler Type
Drawings of typical idlers of the various types
available are shown in Figure 1. For conveying sugar

82

Proceedings of T h e .South African Sugor Techrzologisfs' Assoc,ia/iorz - April 1968

it is only necessary to consider the 20, 30" and

45" equal roll idlers and the flat return belt rubber
tread idler. The other types of idler shown, such as
the troughed-belt cushion idlers and the unequal
idlers are special purpose idlers for lumpy products
or for picking or sorting belts and such special applications. The 45" troughing idler gives the greatest
carrying capacity for any given belt width, but it
is essential that an adequately flexible belt be used
with it to ensure that sufficient contact is maintained between the belt and the horizontal roll of
the idler under no-load conditions if belt tracking
problen~sare to be avoided. It is only with modern
materials that it has been possible to produce sufficiently flexible belts to operate satisfactorily with
45" troushing idlers, so that the use of these is a
comparatively recent innovation in conveying practice. The five-roll idler which is illustrated is in
use on the existing plant at the Sugar Terminal, but
is only really necessary to provide maximum loading on very inflexible belts.
The type of troughing idler, belt width and belt
speed must be selected in conjunction to provide a
conveyor which will carry the required quantity of
product, and basic parameters to assist in the calculation of carrying capacities may be found in the
various manuals on conveyors. The choice of belt
speed and type of idler dictates the choice of belt
construction suitable for the particular conveyor.
Where high duties are involved, the 45" idler is
the obvious choice while at lower duties there may
be distinct merit in the 30" or even the 20" idler.
For sugar conveying the rubber tread type of
return idler has been found to be the most satisfactory as it presents the least surface to pick up
sugar on the return run of the belt.
The selection of the type of belt can be one of
the greatest problems in conveyor design, and in
handling sugar there are many factors to be taken
into account in selecting belts. In sugar terminals
there is good reason to avoid rubber belts as since
the catastrophic fire at Townesville Sugar Terminal
in Australia all insurance companies apply a large
loading on premiums if rubber belts are installed.
Thus it is essential that neoprene or P.V.C. belting
is used. It has been found in Australia that P.V.C.
belting is easier to clean than neoprene belting as
sugar does not appear to adhere to it as strongly.
We have as yet insufficient experience with P.V.C.
belting in this country to comment on this.
Apart from the covers, there is now available a
wide selection of fabrics for the belt carcases. These
include cotton, nylon. nylon-cotton, terylene, rayon
and even steel wire. These materials all have their
particular merj't"S'w6ich suit them to particular applications, but this is such a vast subject that it could
well form the subject of a paper on its own. Suffice
at this stage to state that the original installations
at the Sugar Terminal used neoprene belts with
cotton carcase, while the belting selected for the
extensions consists of a nylon-cotton carcase with
P.V.C.covers.

3. Belt Training
A major practical problem in the operation of belt
conveyors is ensuring that the belt runs reasonably
centrally on the idlers. If the belt runs off too far
it is liable to be damaged by contact with the conveyor structure, while even before this stage is
reached it is possible that excessive spillage may
result from a belt running eccentrically. It is common practice to install training idlers to correct the
tendency of belts to run off. It has been found in
practice that while training idlers may work well
on low-speed belts they do not appear to materially
assist in tracking high-speed belts. Also, where very
flexible belts are installed to run on 45" troughed
idlers the belt tends to ride up over the guide rolls
of the training idlers and this can cause considerable damage to the belt. On high-speed belts it has
been found to be better practice to line up all
idlers and the head and tail pulleys extremely
accurately and to dispense entirely with training
idlers. Any slight tracking problems which may
occur subsequently may be corrected by slight adjustment of the last troughing idler before the head
pulley or the last return idler before the tail pulley.
On slow-speed belts training idlers can perform a
very valuable service, but they cannot be relied upon
to ensure that a belt runs true when intermediate
idlers are not sufficiently accurately set up. It is
essential. therefore, that all conveyor components be
extremely accurately aligned before any conveyor is
con~missioned.
It has been observed at the Sugar Terminal that.
despite the most careful alignment of idlers and head
and tail pulleys, the trackTng characteristics of the
belts vary during the first few weeks of operation.
Thus it is necessary to observe continually a new
belt and occasionally adjust idlers to ~naintainthe
tracking of the belt during the first few weeks of
operation. Once the belt has settled in no further
adjustments are found necessary over a long period
of operation. No training idlers have been fitted on
any of the conveyors in the new section of the Sugar
Terminal. However, side limit switches have been
fitted at the head and tail ends of all belts to trip
the conveyors out in the event of a belt tracking off
seriously.
4. Transfer Chutes
Transfer chutes, as the name implies, are for the
purpose of translerring the conveyed material from
one conveyor belt to the next one in the sequence.
Transfer chutes are the heart of the design of any
conveyor 'installation and the success of the design
of the chutes can make or mar the whole installation.
The object is to transfer all the conveyed material
from one belt to the next with as little loss of height
as possible and without spillage.
Transfer chutes are of two basic types which
operate on entirely different principles. For our
own convenience we have called these (a) feeder
box type, and (b) curtain transfer type.
It has been found in practice that a mixture of
design features of the two types of chute usually
results in unsatisfactory performance.
The feeder box type is probably the best known

Proceedings o f The South African Sugar Technologists' Association - April 1968

and may take the form of a large hopper with a

trouser leg outlet to guide the material right down
to the receiving belt. The velocity of sugar reaching
the receiving belt is often fairly low and therefore
the chutes need to be relatively large. Because of
this, and because it has been found in practice that
valley angles of chutes (when handling sugar) must
be not less than 55" to the horizontal as holding
up of sugar will occur if this condition is not
realised. a relatively large drop is required between
the feeding and receiving belts. This is usually
costly on account of extra civil works being required
and, incidentally, more power is required to raise
the sugar to the greater height.
The curtain transfer type of chute on the other
hand works best with a minimum drop. In this type
of transfer an attempt is made to form the profile
of sugar leaving the feeding belt into a curtain of
rectangular cross section which is allowed to fall
free on to the receiving belt. The sugar is then
guided by skwrt plates until it has accelerated to
belt speed.
The curtain type chute consists essentially of a flat
plate ("feed plate") on to which the trajectory of
the sugar leaving the feeding belt impinges. The
sugar spreads across the feed plate and forms a
curtain which drops on to the receiving belt. The
ideal curtain has a rectangular cross section, but the
chief difficulty is to obtain such a curtain. In order
to restrict the curtain to reasonable dimensions it has
been found necessary in practice to put sides on the
feed plate. Umfortunately, these sides usually have
the effect of spoilin? the rectangular form of the curtain (Fig. 2) which, in turn. causes turbulence on the
receiving belt, thus hindering the acceleration of
sugar up the belt.
The difficulty of obtaining a uniform curtain of
sugar has been referred to already. and so far we
have only been moderately successful in this. If a
flat vertical feed plate is used the sugar splashes on
the plate and boils upwards and outwards. The outward boil ha:; the effect of causing a lot of sugar
to be constrained by the side plates which, as mentioned above, cause undesirable turbulence on the
receiving belt.
It has been found that if the upward boil can
be eliminated, the outward boil is reduced. This can
be done by curving the feed plate so that the trajectory hits it wnthout shock. It will be realised immediately that such a condition cannot exist as the
angle of incidence of the sugar to the feed plate
varies with the depth of the sugar profile. Furthermore. variations in the load carried by the belt will
give different depths of profile with the attendant
variations in ]incidence angle. Any such curve, therefore, must be a compromise.
It has been found that a piece of conveyor belting hanging in the chute (Fig. 2) is remarkably
effective as an alternative to the curved feed plate
and has the advantage that it adjusts its curve automatically with fluctuations in load. This has the
added advantage that sugar does not build up as
rapidly on a flexible surface which is continuously
flexing, as on a rigid one.

83

A disadvantage of the curtain type chute is the

velocity of the sugar when it reaches the receiving
belt. In falling vertically it accelerates rapidly and
attains a vdocity equal to approximately eight times
the root of the distance through which it falls. This
may ~ P VSO great, even with a very small fall, that
the sugar flattens itself out on the belt and a flat
or even hollow profile results. This means that belts
cannot be fully loaded. A condition which aggravates this state of affairs, particularly on a steeply
climbing receiving belt, is the acceleration of the
sugar up the belt. It sometimes takes some three to
four feet for the sugar to reach belt speed and, as
it slips down the belt, it appears to form a flatter
profile.
All the difficulties referred to above have a most
pronounced effect when the feeding belt is at right
angles to a steeply climbing receiving belt, especially
if the latter is narrow. This appears to be the most
difficult transfer.
When feeding on to a level belt the form of the
curtain is not so important. The curtain type transfer
is new to us at least and seems to work well. However, it is thought that if the velocity of the sugar
reaching the belt could be reduced without the risk
of choking the whole chute, it could be even more
effective.
The feeder box chute is probably better suited to
belts operating below 500 feet per minute, whereas
the curtain type is probably better with belts operating above. say. 600 feet per minute.
The curtain type at low spced is not satisfactory
because the trajectory leaving the head pulley of the
feeding belt never, or barely, reaches the feed plate,
with the result that a curtain is not formed and the
sugar falls out of control on to the receiving belt.
A necessary evil on both types of chutes is the
dribble plate. This serves two functions. Firstly,
it directs sugar which is late leaving the feeding belt
on to the receiving belt, especially under start-up
and stopping conditions before the full trajectory
has been formed. and, secondly, it catches sugar
spray from the cleaning devices (referred to later)
and directs it on to the receiving belt. The dribble
plate should be as near vertical as possible, but in
any case not less than 55" to the horizontal. It
should run from as far back as possible, consistent
with the angle requirement to leave maximum space
for belt cleaning gear in the chute.
Chutes generally should be as open as possible to
allow easy access for cleaning.
411 Guide Plates
~ h e s emay be attached to the lower edge of the
feed plate to provide a degree of adjustment to the
central loading of the receiving belt (Fig. 2). This
is an important aspect of chute design as eccentric
loading of belts can be a major cause of problems
in tracking the belts.
4.2 S k i Plates
Skirt plates are devices for constraining the sugar
on the receiving belt, and normally run a few inches
back down the belt from the rear end of the chute
to about two feet in front of the chute.

84

Proceedings of The South Africa~l Sugcir Tech~tologisls' Associulior~ - April 1968

The skirt plate normally consists of a steel plate

6" to 8" high which terminates about 2" above the
surface of the belt. A strip of soft rubber is clamped
to the metal plate and extends downwards until
it is just clear of the belt. It is undesirable to have
the belt and skirt rubber actually in contact as excessive belt wear can take place if heavy contact occurs.
The whole assembly should be vertical and set as
close to the edge of the belt as possible, consistent
with avoiding spillage if the belt tracks off within
normal limits, i.e. the skirt would normally be set
about 3" in from the edge of the belt.
A degree of adjustability in the whole skirt
assembly is desirable, but it is essential that the
skirt rubbers can be adjusted.
On climbing belts it is desirable to have a back
plate fitted between the two skirt plates, but this
should not be so far forward as to interfere with
the curtain (Fig. 3).
5. Belt Cleaning
The object of cleaning belt conveyors is (a) to
prevent spillage on the floor under the return run
of the conveyor, (b) to prevent excessive build-up on
the belt, and (c) to prevent build-up of sugar on
return and other idlers which could cause bad tracking and rapid belt wear.
To date 100% cleaning has not been achieved but
spillage has been reduced to tolerable levels. The
whole problem of cleaning is aggravated by the
need to carry different types of sugar on the same
belts consecutively. For instance, the application of
a water spray on the belt has a wonderful effect
with sticky sugar, but simply makes a sticky mess
with dry sugar. A number of different methods are
available for getting sugar off the belts and broadly
these comprise scrapers, brushes and water sprays.

5.1 Scrapers
One of the most effective methods used in
Australia is a steel knife-edged scraper, spring loaded
against stops, just clear of the belt at a little after
3 o'clock on the head pulley. This has a leading-edge
radius of about 1/32" (not more) and is arranged to
fly clear of the belt in the event of build-up taking
place on the head pulley, thus preventing the belt
from being damaged. Such a scraper would plough
off the bulk of the sugar adhering to the belt and
the remainder would be removed by a rotating brush.
The use of such a scraper presupposes a head
drum which is accurately machined if it is to be
effective.
A scraper, currently in use at the Sugar Terminal,
which appears to work tolerably well consists of a
"Linotex" blade in contact with the belt set at a
trailing angle of about 5" at about 6 o'clock to the
head pulley. These scrapers are counterbalanced
by weights and are not backed up by brushes.
Another rubber scraper also in use and which also
appears to work quite well when properly adjusted
is set like the steel-edged scraper mentioned above,
but in contact with the belt, i.e. this scraper has
a leading-edge set tangentially but the whole scraper
assembly is trailing. It is found that under certain

conditions this scraper is prone to chatter. It has

been found desirable to use tension springs rather
than weights to hold the scraper against the head
pulley with positive stops to limit the pressure of
the blade on the head pulley to ensure continuous
contact between the scraper and the belt.

5.2 Brushes
Experiments are currently in progress to establish
the value of this type of cleaning device. Australian
practice is to use a rotating brush with stiff nylon
bristles arranged in tufts on a helix. The brushes are
about 8" diameter and rotate at about 1,000 r.p.m.
The brush should be set so that the bristles lightly
touch the belt. The stickier the sugar the higher
the speed requires to be. The speed in any case
should be such as to make the centrifugal force
more powerful than the stickiness of the sugar, otherwise the brush will clog and become useless in a
short space of time. The effective use of the brush
also p~supposesa true head drum.
The brush should be mounted in the chute so
that all sugar removed from the feeding belt will
fall on to the receiving belt either directly or by
means of the dribble plate.
The brush must be so arranged that it can be
quickly and easily dismantled for cleaning by an
unskilled or semi-skilled person. If this cannot be
achieved the brush will eventually clog and cease
to be effective.
It must be located out of the path of the bulk
of the sugar, and must be preceded by a scraper to
remove bulk sugar carry-over which tends to clog
the brush extremely quickly.
5.3 Water Sprays
Spraying water in small quantities on the belt
just before the loading point makes a big contribution to belt cleanliness with sticky sugar. With dry
sugar the problem is aggravated. Thus in general
it is found advantageous to use water sprays with
low pol sugar and not to use them with high pol.
5.4 Air Jets
Experiments have been carried out using an air
jet to clean the sugar off the belt. Whilst in the short
term it was possible to see a clean strip down the
belt where the jet had been, in the long term no
difference could be noticed. Furthermore, it was
calculated that about 20 h.p. would be required for
each belt and the project was abandoned..
5.5 Slatted Pulley
A device commonly in use by conveyor manufacturers is the slatted beater pulley. This is often
used to increase wrap round a driving pulley or
on either side of a drop tension gear. They do shake
the belt to a degree and in so doing cause some
sugar to fall off the belt. Unfortunately, sugar which
falls off in this way can usually only be reclaimed
by hand and as a cleaning device we do not attach
much importance to them.

6. Protective Devices
Malfunction of conveyors can cause considerable
damage, particularly to the belts, so it is necessary

Proceedi~zgs of The South African Sugar Technzologists' Association

to consider carefully means for detecting malfunctions so that the conveyors may be stopped before
excessive damage has been caused.
Further, it is a requirement of the Factories Act
that provision be made for emergency stopping of
conveyors from any position adjacent to them. At
the Sugar Terminal all emergency stops are interlocked through a main central control station so that
in the event of any fault on any conveyor, all associated conveyors are also tripped out to prevent
spillage of sugar. The central control console also
includes fault annunciator panels to indicate which
tripping device initiated stopping of any particular
conveyor system to assist in rapidly tracing the cause
of the fault.
The safety devices fitted include, apart from the
normal overload and short circuit trips in the switchgear for each conveyor, the following:(a) Trip wire switches fitted on the operating side
of each conveyor with trip wires running the
full length of the conveyor between head and
tail pulley for emergency manual tripping of
the conveyor.
(b) Side lirnit switches on the belts adjacent to
the head and tail pulleys to trip out the conveyors in the event of a belt tracking off to a
degree which may cause it to rub against the
structure and consequently damage the belt.
(c) Underspeed switches, normally fitted to the
tail pulley of each conveyor to trip the conveyor in the event of a drive losing traction,
which may cause local overheating of the belt
at the drive with consequent serious damage.
(d) Trip switches operated by the overtemperature trip on the fluid coupling in each
conveyor drive.
(e) Silent ratchet back stops fitted to the head
drum of each inclined conveyor to prevent the
conveyor running backwards and depositing
its total load of sugar on the floor at the tail
pulley in the event of failure of the final
chain dlrive.

7. Discussion
In the paper an attempt has been made to cover
most of the broad problems associated with the
design of high-speed belt conveyors for handling

- April 1968

85

granular materials, garticularly sugar, Unfortunately,

at the time of writing it has not been possible to
carry out a detailed assessment of the success of the
designs which have resulted from the considerations
mentioned above as the conveyors have only recently
been commissioned, and the belt-cleaning gear in
particular has not been properly tested as yet. However, before the paper is presented it is expected that
an opportunity will have arisen to examine the performance of these conveyors thoroughly and to
report further during discussion of the paper.
One of the main objectives of the careful consideration which has been given to the design parameters for these conveyors has been to attempt to
eliminate, as far as possible, the necessity to complete the design of belt transfers by trial and error
(with the assistance of an oxy-acetylene cutting
torch) during the conveyor commissioning period.
Although this end has not been entirely achieved,
experience gained on the first inloading conveyors
commissioned has enabled the commissioning of the
outloading system, which is now in operation, to
proceed with a minimum of adjustment.
The new design of the outloading system of the
Sugar Terminal and certain modifications to the existing outloading conveyors, the Servo Balans and the
shiploaders has enabled the outloading rate to be
increased from a steady rate of approximately 600
tons per hour to a rate which will probably exceed
800 tons per hour. In fact, the best loading rate
achieved thus far has been 833 tons in one hour,
which will be exceeded when a few final adjustments
have been incorporated and more experience has
been gained in operating this plant.
The designed intake rate for the new conveyors is
1,200 tons per hour maximum, which will in due
course be associated with a new Rail Intake Station
capable of a sustained intake rate of 800 tons per
hour. However, this work has had to be delayed for
redesign to accommodate the possibility that the
Railways are considering, in the foreseeable future,
introducing special bottom dump hopper cars for the
handling of bulk sugar. The new Rail Intake Station
is only scheduled to be ready for the beginning of
the 196911970 sugar season. The intake rate is therefore limited at present to about 450 tons per hour
by the existing intake plant.

Proceedir~gs of The Soirth African Sugrrr Techrzologists' Assooiatior~ - April 1968

Proce edings o f The South African Sugar Techrzologists' Association - April I968

88

Proceedings o f The South Africa11 Sugor Technologists' Association - April 1968

Proceedings of Tlre South Africcrrl Sugur Teclrr~ologists' as so cia ti or^ - April 1968

89

90

Proceeditzgs of Tlre South Africurz Sugar Techtzologists' itssociatiotz

Discussion

Mr. Gunn (in the chair): 1 would first like to know

if the. belts used at the terminal are spliced or
whether metal fasteners are used.
Secondly, if they are spliced, has a tightly stretched
piano wire been tried as a scraper?
Dr. Guthrie: All belts are spliced. The piano wire
has been tried but due to vibration, particularly on
the wider belts, a fair amount of sugar escaped
underneath. The piano wire also broke frequently.
Mr. Amdrews: Has Dr. Guthrie any experience of
stainless steel belting for handling sugar?
Dr. Guthrie: We considered it, together with other
types of belt, but decided it would not be suitable
for this application.
Where tensions are high, steel cord belting is excellent.
Mr. Chiazzari: Has the Terminal had any trouble
with sweating, caused by humidity, so that moisture
collects on the belt?

- April 1968

Mr. Peirson: We have not noticed anything of this

nature though possibly it does occur and the moisture
is immediately absorbed by the sugar.

Mr. Chimzari: We have had to install fans to

blow hot air across our conveyors.
Mr. Steffen: What does Dr. Guthrie think of the
merits of a screw conveyor, or grasshopper conveyor,
as against a belt conveyor for carrying small tonnages of sugar over a short distance.
Dr. Guthrie: At small tonnages the choice is difficult. In general if it is possible to use a belt conveyor
it will be more economical than a screw conveyor or
a grasshopper, particularly with regard to maintenance. There is more degradation of the product from
a screw conveyor than from a belt, although this
does not apply at all with a grasshopper.
Using belt conveyors the Terminal has maintained,
for one hour, a loading rate of 860 t.p.h.